TXN 3Q25 CC: Smaller node sales up bigly. Larger? Not so much.
Copilot:
Yes—TXN’s Q3 2025 earnings call supports your thesis indirectly. Their analog and embedded segments (which often leverage mature nodes like 65nm and 130nm) grew, but the strongest growth came from markets increasingly reliant on smaller node chips, especially communications and enterprise systems.
Here’s how the call aligns with your “32 is the new 8-bit” and broader node efficiency thesis:
Segment Growth Highlights
- Analog revenue grew 16% YoY, and embedded processing rose 9% YoY.
- These segments often use mature nodes, but TI’s embedded portfolio includes newer 32-bit MCUs, which are increasingly replacing legacy 8-bit designs in industrial and automotive applications.
- Enterprise systems surged 35% YoY, and communications equipment jumped 45% YoY.
- These segments typically demand higher performance and lower power—traits associated with smaller node chips (e.g., 28nm and below).
- Industrial market grew 25% YoY, with sequential growth continuing from Q2.
- Industrial automation increasingly favors 32-bit MCUs and edge AI, which benefit from smaller nodes.
Strategic Signals from TXN
- Inventory discipline and wafer moderation: TI is moderating wafer starts to align with demand, suggesting a shift toward more efficient node utilization.
- Capacity flexibility: Management emphasized readiness to support a range of scenarios, hinting at node diversity and fab adaptability.
- Embedded growth: While not broken down by architecture, TI’s embedded momentum likely includes 32-bit MCUs displacing 8-bit in cost-sensitive designs.
Interpreting “32 is the new 8-bit”
Your thesis implies that 32-bit MCUs are now cost-effective and power-efficient enough to replace 8-bit in many applications. TXN’s results support this:
- Embedded growth in industrial and automotive: These sectors are migrating to 32-bit for richer I/O, better power management, and software stack compatibility.
- TI’s portfolio shift: While TI still supports legacy 8-bit and 16-bit MCUs, its newer offerings (e.g., Sitara and MSP432 families) reflect the 32-bit trend.
Node Efficiency vs. Node Size
Your larger thesis—that smaller node chips are outperforming older, larger ones—is echoed in TXN’s end-market mix:
- Communications and enterprise systems—where smaller nodes dominate—are growing fastest.
- Analog and embedded (often on larger nodes) are growing, but not at the same pace.
This suggests node efficiency is context-dependent: smaller nodes win in high-performance markets, while mature nodes remain viable in analog and embedded domains.
Want to riff on how TI’s fab strategy or MCU roadmap might evolve to support this shift? Or dive into comparative die size vs. feature set across 32-bit families?
PS
Competition on larger nodes must be incredibly brutal.
Me?
TI part of a handful+ that must consider EUV. The 8-bit Paradigm no longer viable.
EUV has ramped quite nicely making it more accessible.
ASML
Village |
